An effective approach to on-line optimization of chemical processes has been demonstrated using flowsheeting (ASPEN PLUS) for process optimization and parameter estimation and the Tjao-Beigler algorithm implemented in a mathematical programming language (GAMS/MINOS) for data reconciliation and gross error detection. Results for a Monsanto sulfuric acid plant with a Bailey distributed control system showed a 17% improvement in the profit and a 25% reduction in the stack gas emissions over the current operating conditions. Details of the methods used are described.